This grant is a competitive renewal of prior 5-year grant, in which it was proposed that fMRI during a memory challenge, in combination with information regarding the presence versus absence of APOE-4, would predict future cognitive decline in the elderly. Studies from the prior funding period did indeed show different patterns of brain activation between the two sub-populations of elderly subjects, in areas known to be involved in Alzheimer's disease (hippocampus, parietal cortex, and dorsal prefrontal cortex); further, increased fMRI activity at baseline correlated with cognitive decline measured 2 years later. The renewal application seeks to expand on these findings by conducting longitudinal follow-up of the genotyped-subject groups, with functional and structural MRI, as well as neurocognitive assessments. New subjects will be added to the sample. Novel fMRI memory activation paradigms will be used and new cortical segmentation and unfolding techniques will be used. The goals of this project are to: define the natural history of changes in memory activation for the two subpopulations of elderly individuals (those with and without an APOE-4 allele); predict long-term outcome from baseline fMRI data; and delineate with greater precision the changes in function and structure of the hippocampus that distinguish elderly with an AD risk factor (APOE-4) from elderly without this AD risk factor.